Debottlenecking process plant

Document Sample
Debottlenecking process plant Powered By Docstoc
					engineering basics

Debottlenecking process plant
In the second of our engineering basics series, Jimmy Hunter explains how
finding and sorting out the problem areas of a process can save money and
improve productivity

          ORPORATE downsizing’ and                information can be a major task.
          ‘sweating the assets’ may be                Another problem is
          familiar terms these days, but in       superstition. Anecdotal
the capital-intensive chemical industry           information from operating and
reducing the payroll is seldom the best           engineering staff can be a rich
way forward. Increasing the output of             source of information, but has to
the existing plant – debottlenecking – is         be carefully filtered. Often
a more powerful way to improve                    ‘everyone knows’ what is causing
productivity, and hence the bottom line.          the bottleneck, but sometimes the
    The word debottlenecking does not             real bottleneck is further up- or
appear in many dictionaries but has               down-stream.
survived because it does its work well.               The chemical engineer has to
When a single component restricts the             use a combination of the available
output of the whole system, we call it a          data, calculations, modelling and
bottleneck. Anyone who has driven on              trial results to get at the truth.
the M6 towards Birmingham will                        Engineering the solution is the     v Bottlenecks limit productivity – wherever they are
understand the concept well.                      most creative part of the process,
Debottlenecking allows the system to              a combination of art and science.
achieve its true potential.                       This may not be the equivalent of                 trials. Using this approach, we can often
    The debottlenecking process seems             painting Leonardo da Vinci’s Mona Lisa,           increase the capacity of the existing
simple. We just have to find the                  but sometimes could come close to the             plant with relatively small modifications,
bottlenecks, engineer the solutions, and          creativity of his helicopter design.              involving low capital cost in a short
successfully implement them. However,                 There are many possible engineering           implementation time.
there are many difficulties and pitfalls.         solutions for debottlenecking a plant.                The final stage of debottlenecking –
Superstition, resistance to change, low           The ‘sledgehammer’ approach –                     implementation – is full of pitfalls. One
job mobility among process engineers              duplicating the entire line – is the least        problem is the attitude of the plant
and our education and training system             risky but very expensive. The ‘partial            personnel. Often the cleverer the
all conspire against successful                   sledgehammer’ – replacing some                    solution the bigger the resistance to it.
debottlenecking.                                  existing equipment with larger versions               This is because of resistance to
    Finding the real bottleneck can be            of the same type – is slightly more risky         change and vested interests. If the
difficult. One problem is that there may          (since scale-up is not always easy or             original plant designer is still around,
not be enough information available.              even possible in some cases), but less            he/she may be hostile because you have
The drive to reduce costs has reduced             expensive.                                        made him/her look bad. The production
the number of instruments that could                  The true afficionado scorns these             personnel may resist the change because
provide clues to the process engineer.            crude techniques and uses them only as            they don’t believe that the solution to
Paradoxically, there is also often too            a last resort. The big savings are made           the problem could be so easy, and
much information, since modern                    using a combination of chemical                   ‘everyone knows that when that flow
computer systems generate huge                    engineering skill and experience,                 reaches 4 m3/h, all hell breaks loose!’
amounts of data. Finding the relevant             computer modelling and well-organised             There is far less resistance to operating a
                                                                                                    new plant at high rates, because there is
                                                                                                    nothing with which to compare it.
                                                                                                        Another problem is dashed expecta-
                                                                                                    tions. The plant and maintenance peo-
                                                                                                    ple have been looking forward to the
                                                                                                    day when they won’t have to crawl
                                                                                                    around under the granulator getting hot
                                                                                                    and dirty. They imagine the shiny new
                                                                                                    plant where everyone sits in the control
                                                                                                    room in spotless overalls, watching the
                                                                                                    computer screens. Your debottlenecking
                                                                                                    efforts will postpone the dream for
                                                                                                    years, and make the area under the
                                                                                                    granulator even more crowded.
                                                                                                        One of the biggest implementation
v Figure 1: Simplified flowsheet of ammonia plant                                                   problems is in the SHE (safety, health

16                                                         the chemical engineer                                                  9 March 2000
                                                                                                          engineering basics

and environment) problems created.
Waiting for the annual shutdown to
make the modifications may be possible,
but is very inefficient. On the other
hand, modifying a live plant ‘on the run’
is a hazardous business and needs very
careful planning and excellent
communication and systems.
    The key to successful implementation
is outstanding teamwork (which implies
good communication) and careful
attention to detail. Careful ‘selling’ of the
debottlenecking project to all the people
concerned, and their involvement in the
implementation from the start, will
ensure success. Sincerely listening to
their problems and concerns, and
dealing with these concerns as part of
the project, will go a long way to getting
their trust and co-operation. Since the
return on capital of a debottlenecking
project can be as much as 2–4 times the
return on a new plant, it is often
possible to include the solution of their
pet hates without jeopardising the              v Figure 2: The original process had four     v Figure 3: The proposed process which
project.                                        wash steps                                    removes the bottleneck
    Careful attention to the ‘battery limits’
– the points at which the modifications
connect to the existing plant – is also         of ammonia as low as possible. The old,       into the reactor did not reduce the
essential.                                      fouled-up cooler was not very effective.      reaction rate.
                                                By redesigning the cooler, the ammonia           These modifications successfully
Example of creative                             recycle to the reactors could be              increased the capacity to 21,000 tpa,
debottlenecking                                 considerably reduced which would              with the casting section as the
An ammonia plant was to be shut down            increase the partial pressure of nitrogen     bottleneck. Developing a mathematical
as soon as its replacement had been             and hydrogen and increase the amount          model of the cooling process in the
commissioned. The new plant project             of fresh feed which could be                  casting operation enabled us to optimise
suffered delays, and the market picked          compressed and converted. The cost of         the operation and expand the capacity
up suddenly. The operators had to keep          the new heat exchanger was relatively         to 24,000 tpa.
the old plant running and squeeze out at        small and successfully provided the
least 5% more production than its               extra capacity needed.                        So how to start?
nameplate capacity. Figure 1 is a                  A second example is a successful           Our education system, with its early
simplified flowsheet of the plant.              expansion of the output of a batch plant      specialisation, and the training of young
   The compressors were old, and                from its nameplate of 16,000 tpa to           chemical engineers – who often become
needed a lot of maintenance.                    24,000 tpa.                                   expert in production of a single product
Mechanical engineers dominated the                 An operations process chart (OPC)          – seldom produces people with the
management and engineering team.                of the original process is shown in           breadth of knowledge and skill needed
They could see clearly that the way to          Figure 2.                                     to successfully debottleneck processes.
get the extra ammonia capacity was to              Calculating the rates of the individual    Hiring contractors is one solution, since
get the most out of the compressors.            sections showed that the concentration,       they often have broad experience.
Since installing a new compressor was           solids washing, and casting steps were        Another way is to hire a process
not an option, they put a massive effort        the main bottlenecks. Note that four          consultant, to work with the project/
into improving the performance of the           washes were used, and two evaporator          process team. Either way, it is important
existing compressors, at considerable           batches were produced per batch of            that the newcomer works closely with
cost.                                           product.                                      the team, to ensure safety and that
   Looking at the chemical engineering             The reaction can be written:               everyone ‘buys in’ to the solution.
of the process, the system involved the                                                           By contrast with the enormous
reversible reaction:                               A + B              Product + Water         bitterness caused by ‘downsizing’, the
                                                                                              fun of being a member of a successful
      N2 + 3H2                 2NH3             We proposed an alternative flowsheet,         debottlenecking team is fondly
                                                as in Figure 3.                               remembered for a long time afterward. s
By Le Chatelier’s theorem, to maximise             This reduced the number of washes
ammonia production you need to keep             to three, and the number of evaporator        For more information contact Jimmy
the partial pressure of nitrogen and            batches from two to one. Since the            Hunter on email hunters@moggs.
hydrogen as high as possible, and that          reaction is irreversible, recycling product

9 March 2000                                              the chemical engineer                                                        17

Shared By:
Description: Debottlenecking process plant